✦
Summary
Unreal Speech is a text-to-speech API that converts written text into natural-sounding speech at a significantly lower cost than many competitors, claiming to be up to 90% cheaper than services such as ElevenLabs and Play.ht. It is aimed at developers, content creators and businesses integrating realistic voice into applications, websites and multimedia projects.
Key Features
Text-to-speech API built on the Kokoro TTS engine
48 natural voices across 8 languages
Real-time streaming with sub-second synthesis
Word-level timestamps via streaming and API endpoints
Async synthesis tasks for up to 500,000 characters
Up to 10 hours of audio per API call
Free tier with 250,000 characters per month; paid plans from $4.99 per month
Volume discounts with unused paid characters rolling over
About Unrealspeech
The API is built on the Kokoro TTS engine and offers 48 voices across 8 languages, including US and UK English, Mandarin, Hindi, Spanish, Portuguese, Japanese, French and Italian. Endpoints cover short streaming synthesis up to 1,000 characters, standard synthesis up to 3,000 characters with timestamp URLs, async synthesis tasks up to 500,000 characters, and a websocket stream with word-level timing, with up to 10 hours of audio per call.
Unreal Speech offers a free tier with 250,000 characters per month, roughly six hours of audio, with commercial use requiring attribution, and paid plans from $4.99 per month with volume discounts and unused characters rolling over. It does not currently offer custom voice cloning.